Hello there,

It occurred to me after reading several replies that the word 'Gain' seems misleading to some while not others.

For a gain less than 1, some people feel that they must call this an attenuation not a gain. I find that, like a lot of other words and ideas, it depends on the context.

For example, in control theory, everything is a gain in the math, and nothing is an attenuation. That's because it's simpler to lump everything into one variable, and certain usages would conflict anyway.

If we are dealing with an amplifier and we know it is an amplifier, we might call the input to output levels a 'gain', and we actual do that. But what if the output is less than the input, do we really want to say this is an 'attenuation' in every case? I don't think so, and that is because if we know it is an amplifier and we assume that amplifiers have a 'gain' then we would still have to call that a 'gain'. If we did not do that, what would we call it then, and amplifier sometimes and an attenuator at other times?

In the math it is much simpler to refer to everything as a gain G:
Vout=G*Vin

or if you rather call it an "amplification factor" then:
Vout=A*Vin

In control theory though it's always a gain, and the gain is often less than 1 because the feedback might sample the output with a resistive divider which effectively divides the output by a constant greater than 1:
Vfb=Vout/K

but we don't usually render it that way, instead we use:
Vfb=Vout*K
where K is a constant less than 1.

One of the reasons for this is because we might have other 'gains' and we don't want to be forced to use gains and attenuations in the same formula, it's simpler and more clear to just use gains:
Vout=G1*G2*Vin

where either G1 or G2 or both can be either greater than 1 or less than 1 or even 1, and we can then lump G=G1*G2 if we wanted to.
If we break it up into gains Kg and attenuations Ka then we might see:
Vout=Kg*Ka*Vint

but then when we lump them we'd still have K=Kg*Ka.

To add more credence to this idea, what if we have an amplifier that has a gain of 1. This is very common. A gain of 1 isn't really a 'gain' in the strictest sense, but it's still called a gain. What else could we call it anyway, we can't call it an attenuation that's for sure, yet it still isn't a larger output from a smaller input.

 

Hello there,

It occurred to me after reading several replies that the word 'Gain' seems misleading to some while not others.

For a gain less than 1, some people feel that they must call this an attenuation not a gain. I find that, like a lot of other words and ideas, it depends on the context.

For example, in control theory, everything is a gain in the math, and nothing is an attenuation. That's because it's simpler to lump everything into one variable, and certain usages would conflict anyway.

If we are dealing with an amplifier and we know it is an amplifier, we might call the input to output levels a 'gain', and we actual do that. But what if the output is less than the input, do we really want to say this is an 'attenuation' in every case? I don't think so, and that is because if we know it is an amplifier and we assume that amplifiers have a 'gain' then we would still have to call that a 'gain'. If we did not do that, what would we call it then, and amplifier sometimes and an attenuator at other times?

In the math it is much simpler to refer to everything as a gain G:
Vout=G*Vin

or if you rather call it an "amplification factor" then:
Vout=A*Vin

In control theory though it's always a gain, and the gain is often less than 1 because the feedback might sample the output with a resistive divider which effectively divides the output by a constant greater than 1:
Vfb=Vout/K

This is an attenuation of K, regardless of whether K is greater than or equal to 1 and whether it is positive or negative.

but we don't usually render it that way, instead we use:
Vfb=Vout*K
where K is a constant less than 1.

This is a gain K, regardless of whether K is greater than or equal to 1 and whether it is positive or negative.

You can describe it either way, but note that K is not the same in these two examples, and thus you should not use K for both -- that is where you cause confusion. Call one Fred and the other Sue. Then it is trivial to establish that, if these equations are for the same system (or part of a system), that Fred = 1/Sue and Sue = 1/Fred.

Q.E.D.

 

This is an attenuation of K, regardless of whether K is greater than or equal to 1 and whether it is positive or negative.



This is a gain K, regardless of whether K is greater than or equal to 1 and whether it is positive or negative.

You can describe it either way, but note that K is not the same in these two examples, and thus you should not use K for both -- that is where you cause confusion. Call one Fred and the other Sue. Then it is trivial to establish that, if these equations are for the same system (or part of a system), that Fred = 1/Sue and Sue = 1/Fred.

Q.E.D.

Hi,

I know what they are sometimes "called" but that is what is being discussed.

The variable 'K' is a constant, nothing more, nothing less. K1 and K2 are constants. If I say the output is:
Vout=Ka*Kb*Vin

it does not matter what we call Ka and Kb, the more general idea is that they are both gains because we don't know what they are yet.

The other idea is to call them simply the "transfer function" which is always referred to as a gain.
Vout/Vin=Ka*Kb

and of course we often see something like:
Vout/Vin=A/(B*s+C)

and it would make little sense to call the 'attenuation' (B*s+C)/A,

but even:
Vout/Vin=A/B

Here B is a constant greater than 1 but it's being used to attenuate the signal.

In control theory we always call it a gain. For the resistive divider where R1=R2 the 'gain' is G=1/2 so we multiply the output by G and get the feedback signal to the control circuit. We can call it an attenuation too, but the math is going to require us to specify it as 1/2 not as 2, because all gains are factors not divisors.

The idea is to make Ka, Kb, A, B, C, N, M, K1, K2, etc., have an ambiguous meaning so we don't have to have two definitions when the math is the same for both ideas.

 

Whether it is greater than or less than one is immaterial to what you call it.

If the relationship is

Vout = Vin * K1

Then the gain is K1. Regardless of whether K1 is positive, negative, greater than one, or less than one.

If the relationship is

Vout = Vin / K2

Then the attenuation is K2. Regardless of whether K2 is positive, negative, greater than one, or less than one.

There's NO ambiguity in the definition.

There is only sloppy and improper use by humans, and nothing is going to change that. You will always have to be aware that people will use the wrong terminology and need to be willing to figure out what they meant as opposed to what they said.

This is common in MANY areas involving math (and areas having nothing to do with math).

If a stock is purchased at $100 some time in the past and is now worth $200, many people will tell you that it increased by 200%, when it didn't. It increased by 100%. There is no need to try to define anything -- the existing definitions are just fine. The problem is that people are bad about abiding by them, and that is not going to change.

 

There is only sloppy and improper use by humans, and nothing is going to change that.
You will always have to be aware that people will use the wrong terminology ..........

For my opinion, the last two words in the text as quoted above lead to the question: What is a "wrong terminology"?
Can a terminology really be "right" or "wrong"?
In this context, I like to quote Sergio Franco (Design with Operational Amplifiers..., McGraw-Hill, 1998.) who is using
the term "transconductance gain" for the output-to-input relation of a VCCS (OTA).

In "The Art of Electronics" we can read that the transconductance of the FET is the "natural gain parameter".

Again - can a terminology be correct or wrong?

 

I am so excited, my stocks gained -10% today. Of course a loss of -10% would be better.

 

In some applications a gain of -1 is an inversion not a loss. (op-amps)

 

For my opinion, the last two words in the text as quoted above lead to the question: What is a "wrong terminology"?
Can a terminology really be "right" or "wrong"?
In this context, I like to quote Sergio Franco (Design with Operational Amplifiers..., McGraw-Hill, 1998.) who is using
the term "transconductance gain" for the output-to-input relation of a VCCS (OTA).

In "The Art of Electronics" we can read that the transconductance of the FET is the "natural gain parameter".

Again - can a terminology be correct or wrong?

Yes, it can.

There can be different conventions, which may result in a particular use being correct in one context and not in another. But most incorrect uses are incorrect in all (even remotely reasonable) contexts.

Consider the stock price example mentioned earlier.

If the starting price is $100 and it goes to $110, would it ever be correct to say that the price increased by 110%? No. And you would never expect to hear someone say that and insist that it was correct.

What would the price be if it increased by 100%. Most people would have no problem determining that the new price is $200. But, if given that the price started at $100 and increased to $200, many of those same people would tell you that the price increased by 200% and insist that they are correct. You can usually get them to eventually see where their thinking is going astray, but not always.

You can say that the new stock price is 200% of it's original price.

Or you can say that the new stock price is a 100% increase from it's original price.

Either of those is correct. But saying that the stock prince increased by 200% is incorrect terminology. Period.

 

I am so excited, my stocks gained -10% today. Of course a loss of -10% would be better.

That's actually a very sore point with me right no. A stock that I have been trying to get out of without taking too much of a loss has been teasing me for a week. Until today, when the company announced that they have been bought for cash and are filing bankruptcy. My investment dropped by 50% immediately before I could get out and I ended up way overinvested in it (my own damn fault). This has cost our retirement funds dearly.

 

That's actually a very sore point with me right no. A stock that I have been trying to get out of without taking too much of a loss has been teasing me for a week. Until today, when the company announced that they have been bought for cash and are filing bankruptcy. My investment dropped by 50% immediately before I could get out and I ended up way overinvested in it (my own damn fault). This has cost our retirement funds dearly.

Ouch.

 

That's actually a very sore point with me right no. A stock that I have been trying to get out of without taking too much of a loss has been teasing me for a week. Until today, when the company announced that they have been bought for cash and are filing bankruptcy. My investment dropped by 50% immediately before I could get out and I ended up way overinvested in it (my own damn fault). This has cost our retirement funds dearly.

I guess they didn't see that coming.

 

Whether it is greater than or less than one is immaterial to what you call it.

If the relationship is

Vout = Vin * K1

Then the gain is K1. Regardless of whether K1 is positive, negative, greater than one, or less than one.

If the relationship is

Vout = Vin / K2

Then the attenuation is K2. Regardless of whether K2 is positive, negative, greater than one, or less than one.

There's NO ambiguity in the definition.

There is only sloppy and improper use by humans, and nothing is going to change that. You will always have to be aware that people will use the wrong terminology and need to be willing to figure out what they meant as opposed to what they said.

This is common in MANY areas involving math (and areas having nothing to do with math).

If a stock is purchased at $100 some time in the past and is now worth $200, many people will tell you that it increased by 200%, when it didn't. It increased by 100%. There is no need to try to define anything -- the existing definitions are just fine. The problem is that people are bad about abiding by them, and that is not going to change.

Hello again,

Either you did not grasp my intent, or I did not explain it well enough.
The ambiguity is FORCED, not by accident, not by mistake, not by sloppy wording. Does this make more sense to you now?
It's also not because of an infinite number of monkeys typing on an infinite number of typewriters (ha ha)

 

I am so excited, my stocks gained -10% today. Of course a loss of -10% would be better.

Hi Bob,

For what I am talking about, if you were to use it in that context (probably would not sound right) that's about what you would say:
"My stocks gained -10 percent today", or more accurately: "My stock gain today was -0.1".
But you would not also say "I am so excited" because then you are misleading the value of the stock.

 

For my opinion, the last two words in the text as quoted above lead to the question: What is a "wrong terminology"?
Can a terminology really be "right" or "wrong"?
In this context, I like to quote Sergio Franco (Design with Operational Amplifiers..., McGraw-Hill, 1998.) who is using
the term "transconductance gain" for the output-to-input relation of a VCCS (OTA).

In "The Art of Electronics" we can read that the transconductance of the FET is the "natural gain parameter".

Again - can a terminology be correct or wrong?

Hi,

That's an interesting view on this. If it makes sense to call something by a particular name, it's probably right.
A rose is a flower, is a plant, is a rose.

 

........
But saying that the stock prince increased by 200% is incorrect terminology.

May I ask: Is this example really about the correct use of the term terminology?
Isn't it more a question of the correct application of percentage calculations which is clearly defined?
In other words, a simple mathematical error?

Returning to the original topic of “electronics”:
What about the term “gain of an antenna” (or simply "antenna gain") which does not fit the definition of an input-output relation?

More tan that, there is also a close relationship between the two terms “terminology” and “definition.”
And I think that a definition cannot be ‘right’ or “wrong,” but rather it must be meaningful (must make sense).

EDIT (added later): To WBahn

Perhaps there is a slight misunderstanding between us.
Let me explain: Of course, somebody can use a "wrong terminology" in a certain case. But this does not mean that this specific terminology would be "wrong". Rather, its usage was not correct in that case. However, I admit - a pretty small difference only.....

 

It is not attenuation, it is loss.

Gain vs. Loss commensurate terms. Things can be amplified or attenuated but that results in gain or loss.

Take an amplifier with a limited bandwidth and input a signal that lies outside it. It is an amplifier but the signal will suffer a loss when passing through it.

 

Hi,

That's an interesting view on this. If it makes sense to call something by a particular name, it's probably right.
A rose is a flower, is a plant, is a rose.

My point was that no one would ever call it a gain, even though it is mathematically consistent. I guess it was too subtle.

 

My point was that no one would ever call it a gain, even though it is mathematically consistent. I guess it was too subtle.

Hi,

See, that's not right. This is my whole point. In this case it may depend on how you use it though.
When you say it the way you originally said it, it sounds wrong, but in another presentation using the word 'gain' would cover every base.

STOCK GAINS YEAR 2025
01/25 +0.02
02/25 +0.01
03/25 -0.02
04/25 +0.01

Without forcing the ambiguity:

STOCK YEAR 2025
01/25 0.02 gain
02/25 0.01 gain
03/25 0.02 loss
04/25 0.01 gain

See the difference?
In this last case you have to be careful to state what it was, gain or loss, but in the first case we just need a signed value.

Could you imagine if we had to do that with say current flow or voltage?
I=0.02 amps left to right (or clockwise)
I=0.01 amps left to right
I=0.02 amps right to left (or counterclockwise)
I=0.01 amps left to right

V=0.02 volts above ground
V=0.01 volts above ground
V=0.02 volts below ground
V=0.01 volts above ground

Now I am not saying that we never do this, but it is much more convenient to lump everything into one variable.
x=+1, x=+2, x=-1, x=+1

or we'd have to do something like this:
x=1 to the right of zero
x=2 to the right of zero
x=1 to the left of zero
x=1 to the right of zero

So by calling everything a gain we avoid the messy explanations. This is always done in control theory. Gains of plus or minus, greater than or less that 1 or equal to 1.

It's funny you quoted stock fluctuations because stock exchanges usually quote plus and minus values because it makes the notation shorter. It's not unheard of to say "gain" or "loss", or "increase" and "decrease", but by using a signed value of SOMETHING it makes it more concise.

Another example, imagine we had a dial for an audio amplifier. When we turn it clockwise the volume goes up, when we turn is counterclockwise the volume goes down. What do we call the dial? Even if the volume goes down to less than the input signal, we still call it "Volume".
In some cases we might get a little fancier, such as with bass cut or base boost, but we could call it "bass level" or just "base".
In other cases we have "bass" and "treble", but we could just use "tone".

 

It is not attenuation, it is loss.

Gain vs. Loss commensurate terms. Things can be amplified or attenuated but that results in gain or loss.

Take an amplifier with a limited bandwidth and input a signal that lies outside it. It is an amplifier but the signal will suffer a loss when passing through it.

Hi there,

Yes, and my point was if we just call either one a 'gain' we don't have to worry about if it is an actual gain or an actual loss.
This is done all the time so I can't figure how why this is so hard to understand for some.
Vout=Vin*K

Is K equal to 1, or is it 0.1, or is it -1, or is it -0.2, or perhaps 10 or -12? We don't have to worry about it, just multiply it by Vin and we get the right answer every single time

I brought this up so that people would perhaps become more comfortable with this idea because after all it is used in a lot of fields, and it wasn't me who started it, it was determined long ago.

 

And I brought up my example to show that, in common parlance, different words are often used for a positive vs negative change.

You will also never hear anyone say its -10 degrees hotter today than yesterday, or the hike went up -1000 ft, or prices are up by .75.

These examples, I would argue, are mathematically correct, but incorrect English usage.